Changes how dezalgo works. Still filters zalgo text without breaking languages. Removes zalgo tolerance because it's not really necessary with this system.
This commit changes /login to no longer take command arguments. Instead, upon sending /login, you will enter an interactive "prompt", where your next OOC message will be interpreted as your login attempt (and will not be shown to other users). This prevents a typo from accidentally broadcasting your credentials to the entire area.
- Changes /login to take no command arguments, /login now sets a client state "is_logging_in".
- pktOocChat will now intercept a client's OOC message if they have this client state, and will interpret that as a moderator login.
- Adds the helper function loginAttempt() for handling moderator logins.
This is the worst code I have ever contributed to this project. This commit is the result of several days of attempting to decipher Qt's documentation. This is the best I can write. Qt has broken me. God help us all.
- Allows sending modcalls to a discord webhook, containing the name of the sender, the area, and the reason.
- Adds configurable options to config.ini for enabling/disabling webhooks and specifying the webhook url.
This is so that tests can use the logic without having to call in main as well.
Furthermore, this way, Qt actually recognises it as a library, and makes it easier to import to tests.
Set a hard limit on 30KB that the server is willing to read, over two sequential reads. If the client sends more than 30KB combined, the server will disconnect the client.
- Limits packets to under 16KB
- Adds a configurable option to config.ini for setting the maximum amount of characters in an IC/OOC message
- Limits the size of OOC names to 30 chars
- The client already limits this, but this is an added precaution.
- Limits the length of shownames to 30 chars
- Implements shownames_allowed for areas, for toggling whether shownames are allowed for messages in that area
